Unified Conformalized Multiple Testing with Full Data Efficiency
报告人： 任好洁(上海交通大学)
时间：2026-10-09 14:00-15:00
地点：智华楼王选报告厅-101
Abstract:
Conformalized multiple testing offers a model-free way to control predictive uncertainty in decision-making. Existing methods typically use only part of the available data to build score functions tailored to specific settings. We propose a unified framework that puts data utilisation at the center: it uses all available data—null, alternative, and unlabelled—to construct scores and calibrate p-values through a full permutation strategy. This unified use of all available data significantly improves power by enhancing non-conformity score quality and maximising calibration set size while rigorously controlling the false discovery rate. Crucially, our framework provides a systematic design principle for conformal testing and enables automatic selection of the best conformal procedure among candidates without extra data splitting. Extensive numerical experiments demonstrate that our enhanced methods deliver superior efficiency and adaptability across diverse scenarios.
About the Speaker:
Haojie Ren is a Professor in the School of Mathematical Sciences at Shanghai Jiao Tong University. Ren received a Ph.D. in Statistics from Nankai University in 2018 and was a postdoctoral researcher at Pennsylvania State University before joining Shanghai Jiao Tong University in 2021. Ren’s research focuses on predictive inference, large-scale multiple testing, and anomaly detection. Her work has been published in journals including JASA, Biometrika, JRSSB and JMLR.
Your participation is warmly welcomed!
欢迎扫码关注北大统计科学中心公众号，了解更多讲座信息!
